SENTIMENT ANALYSIS OF APEX LEGENDS GAME REVIEWS ON STEAM USING NAÏVE BAYES CLASSIFIER

Abstract

The rapid development of video games, which can now be played online through several platforms, is remarkable. One of the commonly used platforms is Steam, where each game has reviews, but some reviews are biased and ambiguous. In this case, sentiment analysis is useful to determine whether the reviews contain positive aspects like improving strategic skills or negative aspects like the potential for addiction. It is also used to evaluate the model's performance in identifying these reviews. The sentiment analysis model was created using Word2Vec and Naive Bayes methods. The research resulted in a sentiment analysis model with an accuracy of 75%, precision of 83%, and recall of 82% from 4332 review data, consisting of 3087 positive and 1245 negative data. Therefore, it can be concluded that Apex Legends game is positively received by users.